Obesity Representations: An Analysis of Italian Newspapers’ Coverage before and after COVID-19 and the Influence of Political Orientation

The World Health Organization (WHO) refers to obesity as a “complex multifactorial disease, defined as abnormal or excessive accumulation of fat that presents a risk to health” [1, p. 3]. In recent decades, the number of obese people has increased exponentially from about 100 to 671 million [1], and currently, more than half of the global population (59 percent) is reported to be obese or overweight [2]. As a consequence of this considerable increase, obesity has been defined as an epidemic [3‒5], joining other significant health concerns such as malnutrition and infectious diseases as a leading cause of health risks [5].

During the period of the COVID-19 pandemic, there has been a notable increase in media attention devoted to the topic of obesity. This is largely due to the rise in the number of deaths and intensive therapy hospitalizations of obese individuals [6]. Given the substantial body of literature demonstrating the impact of media representations on people’s attitudes [7‒16], several studies have commenced an investigation into the representation of obesity in the media [17]. Indeed, the manner in which the media represent obesity has frequently been identified as a significant contributing factor to the formation and perpetuation of weight-related stigma and discriminatory practices (e.g., [18‒21]). For example, exposure to news about the obesity epidemic was observed to exacerbate anti-fat prejudice, thereby impeding the promotion of less prejudicial attitudes [22]. A similar trend was observed in British national newspapers which, by portraying obesity and overweight in a caricatured manner and ascribing controllable causes to it, played a pivotal role in shaping and perpetuating anti-obesity attitudes [23].

This appears to be due to the fact that the media typically portray obesity as an issue related to personal responsibility and behavior [24], while overlooking potential broader systemic factors such as junk food marketing, processed food supply, social isolation, and poverty [20, 25]. A number of studies have demonstrated that the prevailing media narrative concerning obesity is primarily focused on health-related issues, the economic burden on the healthcare system and society at large [26], potential causes, and possible solutions [27]. Additionally, there is a tendency to represent obesity as an obstacle to national progress, which can be overcome through a collective effort [28]. More Furthermore, studies employing frame analysis [29] have revealed that the media representation of obesity encompasses a range of discourses, including fatalism [30], self-control [31], war, religion, addiction, and epidemic [17], as well as the personal, interpersonal, and social dimensions of the issue [32, 33].

It is noteworthy that the frames used by the media to represent obesity exhibited different trajectories over time [24]. For example, it has been observed that over the past decade, the US media have shifted from conveying a representation of obesity as due to individual causes to one focused on social causes, in response to a change in the social context [34]. In contrast, a different trend has emerged in the UK, where the media have increasingly conveyed a representation of obesity as a biomedical problem caused by individual actions, and less and less in relation to its environmental determinants [35]. Such changes over time appear to reflect the influence that external events can exert on representations (e.g., [36]). Indeed, legal action taken against fast food outlets has been observed to have contributed to an increased focus on the social causes of obesity [37]. Similarly, the period of the COVID-19 pandemic, which has been characterized by significant mental health distress [38], has resulted in the emergence of a more stigmatizing discourse surrounding obesity. This has involved portraying individuals with obesity in fatalistic terms and as a contributing factor to the problems faced by the country’s health care system [39].

Another factor that appears to impact the representations conveyed by the media is their political-ideological orientation [40, 41]. For example, it was observed that while left-wing newspapers conveyed a representation of obesity that reflected the role of institutions and right-wing newspapers the risks to the individual [39], newspapers of both political orientations focused on individual factors, neglecting political and social factors [24]. Cotter and colleagues [17] conducted a study in which they selected newspapers articles according to their political orientation, but they did not investigate the association between these articles and the frames that emerged. This suggests that the association between representations and political orientation is rarely considered in studies of media representation of obesity. It is worth noting that the aforementioned research employs top-down methods that rely on predefined coding systems to categorize units of text, thereby mapping the representations of obesity. These methods comprise a set of content categories that function as a coding system and a set of rules designed to apply the categories to the text [42]. However, according to a pragmatic, dynamic, and contextual view of meaning [43, 44] distinctive of the Semiotic Cultural Psychology Theory (SCPT [45, 46]), the analysis of a text cannot be carried out through the application of coding rules that are blind to the context (i.e., according to which if word x occurs, content A has arisen). Rather, it must be carried out through an inferential reconstruction of the linguistic and/or extralinguistic context of the text [42]. According to this perspective on meaning, “the word x in the context of its connection with the words y and z means A; but in the context of its connection with the words m and n it means B” [42, p. 257].

To date, one study has focused on mapping the representation of obesity through the analysis of Italian Twitter data [47]. However, there are no studies that have focused on the representations conveyed by the Italian national press. Moreover, existing research appears to lack a more context-sensitive analysis, which could provide a more nuanced understanding of how obesity is represented in the national press and how such representation is influenced by the political orientation of the newspapers. A study that considers these aspects could provide new insights into the representations of obesity conveyed by the Italian press and their evolution over time. It would also help to fill a gap in the literature regarding the role of the political orientation of newspapers in promoting specific representations of obesity.

The present study aims to identify the representations of obesity in nine main Italian national newspapers, analyzing how these representations vary over time and investigating the role of the newspaper’s political orientation in determining the prevalence of specific narratives over others. It is expected that the number of articles dedicated to the topic of obesity will increase in the post-COVID-19 period (H1). Moreover, it is expected that during the period of the COVID-19, specific representations will have a higher incidence than others (H2). Finally, it is expected that newspapers will convey different representations over time, contingent on their political orientation (H3).

The sample selection was conducted according to four steps: the initial selection of newspapers and the delineation of the time interval, the definition of the political orientation of the selected newspapers, the definition of keyword, and the selection of articles. National newspapers that permitted free access to their online archives and the downloading of articles were deemed suitable for consideration. The time interval was selected to encompass the years preceding and succeeding the onset of the COVID-19 pandemic. The prepandemic period was defined according to the restrictions imposed on online archives of certain journals, which did not permit free access to articles published prior to 2016.

To determine the political orientation of the selected newspapers, an online questionnaire was administered via the PsyToolkit platform [48, 49]. The participants were asked whether they read newspapers and what they deemed to be the political orientation of the nine newspapers selected in the present study. Responses from participants who indicated that they did not read newspapers were excluded from the analysis. The SM2 contingency table, which can be found in the online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000542544), illustrates the frequency of participants’ opinions regarding the political orientation (in column) of the selected newspapers (in row). To ensure an optimal distribution of newspapers according to the political orientation, the newspapers were classified into three macrocategories: Center-left, Center, and Center-right also if the participants could choose among five categories (Left, Center-left, Center, Center-right, and Right). The classification was conducted according to the following criteria: (a) a newspaper was assigned to the Center-left macrocategory if the sum of the frequencies of the Left and Center-left categories was greater than both the frequency of the Center category and the sum of the frequencies of the Right and Center-right categories. Similarly, a newspaper was classified as Center-right if the sum of the frequencies of the Right and Center-right categories was greater than both the frequency of the Center category and the sum of the frequencies of the Left and Center-left categories. (b) In cases where the sum of the frequencies of the Left and Center-left or Right and Center-right categories was equal to or lower than the frequency of the Center category, the newspaper was assigned to the Center macrocategory. This decision was made because the Center category was the only one that was not included in the calculation with the additional categories.

To define the keywords to be used in the search of the articles, an already established method was employed [50‒55]. This method involved conducting a preliminary frequency analysis of the words associated with the investigated topic on a sample of newspaper articles. Two independent judges conducted a search of online articles using Google News, each selecting the five articles they considered to be related to the topic of obesity. The word most frequently occurring in each of the 10 articles and semantically related to the topic of obesity was identified as a candidate keyword. Subsequently, each of the two judges employed the candidate keyword to conduct a preliminary online search of an additional five articles. By comparing the selected articles, the two judges, through a consensus procedure [56, 57], determined whether the candidate keyword was able to discriminate articles related to the topic of obesity.

In the preceding step, a keyword was identified and subsequently utilized in conjunction with the word “obesity” for the purpose of selecting articles. This selection was conducted according to the following criteria: each article was required to contain at least one of the aforementioned keywords; the keyword was required to appear a minimum of three times within the article; and each article was required to include a minimum of 250 words. Duplicate articles were excluded from the analysis. Figure 1 provides a visual representation of the aforementioned steps.

This study received ethical approval from the Ethical Committee of the Department of Dynamic and Clinical Psychology, and Health Studies, Sapienza University of Rome (Protocol No. 0001291, July 26, 2022) and adhered to the guidelines outlined in the Declaration of Helsinki, which was adopted by the World Medical Association (WMA) during the 18th WMA General Assembly held in Helsinki, Finland in June 1964. Written informed consent was obtained from all participants.

A textual data analysis was performed following the Automated Co-occurrence Analysis for Semantic Mapping procedure (ACASM [42, 58‒60]). The ACASM is a bottom-up text analysis procedure that has been previously employed in studies examining representations of various topics in the media, including the newspaper coverage of the COVID-19 [53, 61], climate change [54], immigration [51], eating disorders [62], and adoptions [63]. The ACASM is used to map the thematic content (henceforth: themes) active in the text corpus and consists of different automatic steps (cf. online suppl. material for technical details): transcript segmentation, lemmatization, lemma selection and dictionary construction, digital text representation, and cluster identification. In comparison to alternative text analysis methods based on coding (e.g., [64, 65]) – which are time-spending and prone to reliability and replicability issues due to the high degree of inference required for their implementation, or on the use of dictionaries (e.g., [66, 67]) and word counting (e.g., [68]) – which are based on a view of meaning as content defined a priori and independent of the context [69, 70], the ACASM procedure allows mapping the meaning of a text by focusing on word co-occurrences. Accordingly, each pattern of co-occurring words is considered an indicator of an active theme within the text, under the assumption that such a pattern gives rise to a particular semantic content, which the interpreter must activate in order to make sense of that pattern [59].

To assess the statistical processability of the data, the type/token ratio (values less than 0.20 are deemed acceptable) and hapax percentage (values less than 50% are considered adequate for large corpora [>200,000 tokens]) of the corpus were checked [71]. Subsequently, themes were identified through cluster analysis (CA) [72] with a bisecting k-means method [73], which identifies specific patterns of co-occurring lemmas within the same textual unit (i.e., paragraph). The number of clusters to be selected was determined by comparing the Calinski-Harabasz Index [74] (high values indicate better clustering), the Davies-Bouldin Index [75] (low values indicate better clustering), and the intracluster correlation coefficient of the various cluster partitions that emerged from the CA. The complete ACASM procedure was conducted using T-LAB software [76].

Once the optimal partition had been selected, cluster labeling was conducted through a two-step qualitative procedure. First, in order to identify a set of potential labels 10 psychology students from Sapienza University of Rome who were unaware of the research objectives were asked to (a) examine the 10 characteristic lemmas of each cluster and (b) for each cluster provide three labels without repeating the ten characteristic lemmas of the clusters. Second, an additional ten psychology students from the same university, also unaware of the research objectives, evaluated the representativeness of each of the 30 words collected in the previous step using a seven-point Likert scale (from 1 = not at all representative to 7 = totally representative). The word with the highest score was selected as the label for the corresponding cluster.

To test the first hypothesis of the study, each year of the specified time interval was divided into quarters, resulting in a total of 30 distinct time blocks. In consideration of the official declaration of the state of pandemic emergency in Italy on March 31, 2022, the time block corresponding to the first quarter of 2022, corresponding to the time block 25, was established as a demarcation point between the pre- and postpandemic periods. Subsequently, a t test was conducted using Jamovi software (version 2.5.4 [77]) to ascertain whether there were significant differences in the number of articles published between the two periods under consideration. To assess the prevalence of certain representations over others during the COVID-19 period, and to determine whether this prevalence was influenced by the political orientation of the selected newspapers, a time series analysis was conducted using the RStudio software [78] and the packages strucchange [79] and breakpoints [80].

According to the aforementioned criteria (cf., § Sample selection), nine national newspapers were selected: Corriere Della Sera, Il Fatto Quotidiano, Il Foglio, Il Giornale, Libero, Il Mattino, Il Messaggero, La Repubblica, La Stampa. To include the years preceding and succeeding the COVID-19 pandemic, and in light of the limitations on access to the online archives of certain newspapers, the time interval from January 2016 to July 2023 was selected for the article search. The procedure employed to ascertain the political orientation of the newspapers resulted in the classification of Il Fatto Quotidiano, La Repubblica, and La Stampa as Center-left; Il Mattino and Il Messaggero as Center; Corriere della Sera, Il Foglio, Il Giornale, and Libero as Center-right (cf. online suppl. Table SM2 in supplementary material).

The procedure adopted for the identification of keywords revealed that the words “overweight” and “fat” are semantically related to “obesity.” As reported in online supplementary Table SM1 of the supplementary material, the word “overweight” was selected due to its higher frequency in the preliminary articles reviewed. Therefore, the articles were searched using the words “obesity” and “overweight” as keywords. A preliminary search, based on the selected keywords and time interval, yielded 874 articles. According to the established inclusion criteria, 69 articles were excluded due to the absence of at least three occurrences of either of the two keywords, 18 articles were removed for containing less than 250 words, and two articles were excluded as duplicates. Finally, 785 articles that met the inclusion criteria were selected. Table 1 provides a summary of the selected articles from each newspaper, organized according to their political orientation within the specified time interval.

The digital text representation matrix resulting from the automated procedure performed by T-LAB (see online suppl. material) consisted of 9,582 elementary context units (ECUs) in rows and 16,172 lemmas in columns. The digital matrix was found to be statistically processable, as evidenced by a type/token ratio of less than 0.20 (0.06) and a hapax percentage of less than 50% (0.41) considering the corpus used as large (tokens = 424,065).

The CA performed on the digital matrix yielded a classification rate of 99.92% of the ECUs (9,574 out of 9,582 total), with the identification of five clusters as ideal partitions. (Table 2 presents a comparison of the Calinski-Harabasz, Davies-Bouldin indices, and intracluster correlation coefficient of extracted partitions.)

As illustrated in Table 3, the two-step qualitative procedure described above resulted in the following cluster labeling:

• Wellness (21.49%): This cluster groups lemmas that refer to physical activity (physical, activity), the importance of proper eating habits (nutrition, habit, food, correct, style) and the role played by the family (family, parent) and education system (school) in promoting them.

• Problem (28.90%): Lemmas in this cluster deal with the epidemiological aspect of obesity (million, overweight, obesity, population, healthcare) and its incidence in Italy (Italy, Italian), Europe (European), and modiale (Country, World).

• Health (14.35%): This cluster brings together lemmas that relate to the medical aspects associated with obesity (body mass index) and the pharmacological (therapy) and surgical (surgery, bariatric, surgical, operation) treatments (treatment) that patients (patient) undergo.

• Nutrition (20.23%): Lemmas in this cluster relate to the properties of foods that are consumed (food, sugar, eat, drink, fruit, consumption), their effects at the body (fat, cell) and brain (brain) levels, and in laboratory experiments with animals (mouse).

• Disease (15.03%): This cluster contains lemmas that refer to the risk (risk, factor) of developing specific diseases (tumor, diabetes, cancer) and experiencing certain problems (hypertension, cardiovascular, pathology) that the research (study, publish) has shown to be associated with obesity.

Examples of ECUs, and their characteristic words (in bold), associated with each cluster are given below:

• Wellness: Moreover, there is an association between children’s physical inactivity and that of their parents. It is mostly young people whose parents say they do not participate in sports or physical activity who have sedentary lifestyles: 47.9% if both parents are sedentary versus 9.8% if neither parent is sedentary.

• Problem: […] teenage girls increased from 5 million to over 50 million and boys from 6 million to 72 million (a total of 123 million so classified to which 213 million overweight should be added). At the other extreme are malnutrition and underweight children: 75 million girls and adolescents moderately or severely underweight and 117 million boys.

• Health: At Humanitas Gavazzeni, we treat patients with a body mass index above 40 and, in case of the presence of two metabolic comorbidities such as diabetes or hypertension, also patients with a body mass index between 35 and 39. Our patients are mainly women because they are more prone to medical checkups with an age around 36 years: young, therefore.

• Nutrition: The tips to follow are very simple: reduce consumption of packaged foods (high in fat and salt), avoiding snacks and industrial snacks; reduce consumption of carbonated and sugary drinks; and increase consumption of fruits, vegetables, and grains.

• Disease: Type 2 diabetes and obesity are risk factors for cancer occurrence: a study confirms. Type 2 diabetes and obesity are risk factors that predispose to the occurrence of at least eight different cancers: breast, endometrial, colorectal, liver, pancreatic, esophagus, gallbladder, and ovarian.

The t test performed with the number of articles as the dependent variable and pre-/post-COVID-19 as grouping variable showed a significant difference between the two groups (F [28] = −2.43, p = 0.022). Specifically, more articles on obesity were published on average in the post-COVID-19 than in the pre-COVID-19 period (mean difference = −8.68).

The time series analysis showed that some themes had a more significant incidence than others during the period of the COVID-19 pandemic (Fig. 2). Specifically, Wellness was characterized by a significant breakpoint (F = 19.23, p < 0.001) at time block 25. Specifically, there was an increase in articles conveying this representation until the first quarter of 2022. The Health theme had a significant breakpoint at time blocks 20 (F = 23.07, p < 0.001). Specifically, there was an increase in articles conveying this representation in the fourth quarter of 2020. The themes Problem, Nutrition, and Disease showed no significant variations in the selected time interval (F = 6.22, p = 0.202; F = 2.05, p = 0.884; F = 7.01, p = 0.144; respectively). Figure 2 summarizes these results.

Time series analysis, segmented according to the political orientation of newspapers, revealed that the representations of obesity conveyed by Italian newspapers varied according to their political orientation. Specifically, for the Center-Left newspapers, it was found that the Wellness and Health themes showed no significant breakpoints (F = 5.05, p = 0.326, F = 2.20, p = 0.853; respectively). The Problem theme showed a not significant breakpoint (F = 8.82, p = 0.065) at time block 17 (first quarter of 2020). The Nutrition theme showed significant breakpoints at time blocks 16 and 25 (F = 16.08, p < 0.001), with an increase in the fourth quarter of 2019 and a decrease until the first quarter of 2022, respectively. The Disease theme showed a significant breakpoint (F = 14.10, p = 0.005) at time block 19, corresponding to an increase of such theme until the third quarter of 2020.

For the Center newspapers, a significant breakpoint emerged in all the themes. Specifically, Wellness (F = 53.94, p < 0.001) had a breakpoint at time block 23 (an increase until the third quarter of 2021), Problem (F = 13.89, p = 0.006) at time block 25 (an increase until the first quarter of 2022), Health (F = 24.96, p < 0.001) at time block 19 (an increase until the third quarter of 2020), Nutrition (F = 9.60, p = 0.045) at time block 13 (an increase until the first quarter of 2019), and Disease (F = 10.92, p = 0.024) at time block 13 (an increase until the first quarter of 2019).

Finally, for the Center-right newspapers, there was a significant breakpoint in the themes Wellness (F = 15.45, p = 0.003) at time block 24 (an increase until the fourth quarter of 2021), Problem (F = 10.95, p = 0.024) at time block 25 (an increase until the first quarter of 2022), and Health (F = 30.80, p < 0.001) at time block 20 (an increase until the fourth quarter of 2020). The Nutrition theme showed two not significant breakpoints (F = 4.17, p = 0.455) at time blocks 8 and 12 (fourth quarter of 2017 and 2018, respectively). The Disease theme showed a not significant breakpoint (F = 8.15, p = 0.087) at time block 6 (second quarter of 2017). Figure 3 summarizes these results.

This study was aimed at identifying the representations of obesity conveyed by nine main Italian newspapers, examining their incidence and variation over time, and analyzing how the political orientation of newspapers influences the prevalence of specific representations over others. The finding revealed that, within in the Italian context, newspapers represented obesity according to five different themes labeled Wellness, Problem, Health, Nutrition, and Disease. The identified themes are consistent with the content and frames found in previous studies. For example, the Wellness theme – which referred to physical activity, family habits, and a correct lifestyle – has recently emerged as a central aspect promoted by health professionals and on which behavioral interventions aimed at addressing the growing prevalence of obesity are based [81]. The Problem theme – related to the epidemiology of obesity – has previously emerged in the literature [35] and is consistent with studies that addressed obesity in terms of a growing global epidemic [3‒5, 17]. The Health theme – which pertains to the medical aspects associated with obesity (e.g., pharmacological and surgical treatments) – has been previously identified in previous studies that have highlighted the pivotal role of the healthcare system in the management of obesity [35] and how the latter has been framed and discussed in terms of health [82]. The Nutrition theme – which concerns the properties of the consumed foods and their effects on the body and brain – is consistent with studies that demonstrated how the media tended to frame obesity in relation to unhealthy eating behaviors [37, 81, 83]. The Disease theme – which pertains to the risk of developing specific diseases – is consistent with previous studies [17, 35] that showed how the media predominantly framed obesity primarily in relation to the associated diseases [84].

The results showed that following the official declaration on March 31, 2022, concerning the conclusion of the state of emergency related to the COVID-19, there was a significant increase in the number of articles published on the topic of obesity in nine Italian newspapers. This finding substantiates the first hypothesis of the present study. This phenomenon may be attributed to the considerable impact of the COVID-19 pandemic on the prevalence of obesity in the population [85]. This assertion is supported by a study that predicted that COVID-19 pandemic would have long-term negative effects on the obesity epidemic [86]. Indeed, the implementation of lockdown measures has contributed to an increase in sedentary behavior and the emergence of less healthy eating habits [87]. Furthermore, this increase was also observed among school-age children and was directly linked to the disruption of daily routines and increased sedentary behavior during periods of lockdown [88].

The results demonstrated that, during the COVID-19 pandemic, Italian newspapers represented obesity in terms of Wellness – i.e., highlighting aspects related to physical activity, the relevance of proper eating habits, and the supportive role of the family and educational system – and Health – i.e., emphasizing the medical aspects associated with obesity such as the pharmacological and surgical treatments that patients often have to undergo. This interest may have been driven by two factors. First, there was growing awareness of the benefits of a healthy lifestyle in the context of a global health crisis [89]. Second, obese and overweight people were considered a vulnerable group during the pandemic [90]. This was because there was evidence that obesity could accelerate the adverse clinical outcomes of COVID-19 [91].

The results showed that the variation over time of the different representations exhibited distinct patterns according to the political orientation of the newspapers. In particular, Center-left newspapers showed an increase in articles that represented obesity in relation to the properties of foods consumed and their bodily effects (Nutrition), as well as the risks associated with specific diseases (Disease). In contrast, articles representing disease according to the themes of Wellness, Problem, and Health did not exhibit a significant variation. This suggests that Center-left newspapers have maintained a stable narrative focus and editorial line, despite external influences such as the pandemic.

In contrast, Center-right newspapers showed an increase in coverage related to obesity, with a particular emphasis on physical activity, eating habits, and family and educational system support (Wellness), as well as epidemiological (Problem) and medical (Health) aspects. This surge in coverage was particularly pronounced during the pandemic period. Therefore, Center-left newspapers seem to show a greater focus on obesity-related issues in terms of Nutrition and Disease, as early as 2019 (prior the lockdown due to the COVID-19). In contrast, Center-right newspapers exhibited a surge in interest in obesity-related matters in terms of Wellness, Problem, and Health only after the onset of the pandemic lockdown. Such a difference may be indicative of a broader inclination toward the prevention and promotion of healthy lifestyles, which is consistent with the social and health policies supported by the Center-left newspapers. However, the considerable surge in coverage during the pandemic also suggests a reactive stance by the Center-right newspapers, emphasizing social and health concerns that were previously marginalized. This surge in coverage may be an effort to address mounting public concerns about the pandemic and its health implications, as well as a reflection of an evolving editorial agenda in response to shifting public priorities. Coherently, Center newspapers showed an increase in interest for obesity in terms of Health, Nutrition, and Disease prior to the pandemic lockdown and only after it in terms of Wellness and Problem. This may indicate an effort to adapt the narrative to changing public concerns over time. It could also represent an attempt to respond to broader sociopolitical dynamics by trying to balance and integrate different Center-left and Center-right perspectives.

The results of the present study are interesting from an applicative perspective. The capacity to identify the prevailing themes through which newspapers represent obesity and to trace their evolution over time offers a deeper understanding of how the phenomenon could be represented at the general population level. The findings of the present study could be useful in the development of awareness campaigns focused on wellness promotion. Such campaigns may seek to educate the public about the impacts of obesity on psychophysical well-being, adopting a positive approach that emphasizes the benefits of healthy behaviors, such as a balanced diet and the importance of regular physical activity. In this manner, a cultural transformation may be facilitated that encourages a healthy lifestyle, thus contributing to the prevention of obesity itself and the onset of related health conditions. Preventive campaigns could focus on nutrition education, providing comprehensive information about the consequences of poor dietary habits while simultaneously promoting more conscious and healthy choices. Moreover, the reduction of the prevalence of specific narratives that categorize obesity as a problem or a disease could cultivate constructive attitudes and diminish the stigma associated with obese individuals. Indeed, research indicates that when the language used to refer to overweight or obese individuals is more positive, it can help to reduce the prevalence of stigmatizing content [20].

Although the findings of the present study have important implications, it is also important to acknowledge the limitations of the study. This study focused exclusively on the Italian context and considered only the principal newspapers that permitted access to the online archive. As a result, the sample was not perfectly balanced sample in terms of political orientation, which limits the generalizability of the results. Future research should address these limitations by conducting cross-national comparative studies, which would assist in identifying potential variations in representations of obesity in different countries and cultures. Furthermore, the inclusion of a wider range of newspapers in future research would facilitate a more balanced sample in terms of political orientation and a more comprehensive understanding of the representation of obesity. In addition, a comparison of the number of articles related to obesity with the total number of published articles each year for each newspaper could provide a more comprehensive understanding of the attention given to the issue of obesity. From a statistical standpoint, this would permit the data to be normalized and facilitate an assessment of the relative proportion of attention devoted to the topic of obesity, thereby enabling newspapers of different sizes to be compared in terms of editorial productivity. Moreover, this would reduce unexplained variability (e.g., changes in the total number of publications due to editorial, economic, or organizational factors), thereby enhancing the reliability of the data for comparisons across journals and time periods. Additionally, the sample recruited to determine the political orientation of newspapers was small, comprising only 47 participants. The limited sample size may not accurately reflect readers’ perceptions of the political orientation of newspapers. Indeed, some participants classified the newspaper Libero as Left-wing or Center-left, a classification that appears to be at odds with the existing literature, which identifies it as a right-wing or far-right newspaper [92‒94]. It is important to note that assigning a specific political orientation to a newspaper is a challenging endeavor. Indeed, the political orientation of a newspaper may shift as a result of changes in the editorial management team. To overcome the aforementioned complexities, this study adopted a classification methodology that divides newspapers into three macrocategories: Center-left, Center, and Center-right. However, it must be acknowledged that these categories may not fully encompass the complexity and political polarization of some newspapers, particularly those with a more radical orientation. It would be beneficial for future studies to utilize a larger sample size, which would ensure a more comprehensive and representative assessment on the political classification of newspapers. In addition, it would be useful to implement alternative classification methodologies, integrating and comparing the results with existing literature, in order to enhance the accuracy of the political labels utilized.

In conclusion, the present study helps shed light on the representations of obesity in the Italian press, their evolution over time, in relation to the pandemic lockdown, and how these are influenced by the political orientations of the newspapers. The media, policy-makers, health professionals could consider these findings when developing strategies to promote positive narratives, taking into account the associations between media representations of obesity and different social processes, such as stigmatization.

This study received ethical approval from the Ethical Committee of the Department of Dynamic and Clinical Psychology, and Health Studies, Sapienza University of Rome (Protocol No. 0001291, July 26, 2022) and adhered to the guidelines outlined in the Declaration of Helsinki, which was adopted by the World Medical Association (WMA) during the 18th WMA General Assembly held in Helsinki, Finland in June 1964. Written informed consent was obtained from all participants.

The authors declare that they have no conflicts of interest.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conceptualization: M.R. and C.L.; methodology, data curation, and formal analysis: M.R. and G.V.; investigation and visualization: V.C., F.L., and A.L.; writing – original draft, review, and editing: all authors; and supervision and project administration: C.L.

The data that support the findings of this study are available online at: https://osf.io/3jks2/.